get_map_layer_prefix <- function(text){
  return(substring(text, 0, 4))
}

get_map_layer_name <- function(text) {
  return(substring(text, 5))
}


#Server handling user input and processing of data
server <- function(input, output, session) {
  currently_selected_bezirk <- reactiveVal("")
  currently_selected_stadtteil <- reactiveVal("")
  currently_selected_maptype <- reactiveVal("Bezirke")
  currently_selected_heatmap_crime <- reactiveVal("Allgemeine Verstöße gem. § 29 BtMG -Konsumentendelikte-")
  
  
  currently_compared_location <- reactiveVal(c(""))
  currently_compared_crimes <- reactiveVal(c(""))
  currently_compared_year <- reactiveVal(c("2024", "2023"))
  
  # 1. Aktualisieren der Auswahlmöglichkeiten mit den extrahierten Schlüsselnamen
  updateSelectizeInput(
    session = session, 
    inputId = "search", 
    choices = auswahlmöglichkeiten,
    selected = ""
  )
  
  observeEvent(input$search, {
    req(input$search)
    currently_selected_bezirk(get_bezirk_by_stadtteil(input$search))
    currently_selected_stadtteil(input$search)
  })
  observeEvent(input$heatmap, {
    currently_selected_heatmap_crime(input$heatmap)
  })
  
  observeEvent(input$rd_maptype, {
    currently_selected_maptype(input$rd_maptype)
  })
  
  observeEvent(input$vergleichs_jahr, {
    currently_compared_year(input$vergleichs_jahr)
  })
  
  observe({
    maptype <- currently_selected_maptype()
    heatmap <- currently_selected_heatmap_crime()
    mapproxy <- leafletProxy("hhmap")
    clearGroup(mapproxy, "selected")
    clearGroup(mapproxy, "layer_heatmap")
    # Entfernt eine eventuell existierende Legende vom vorherigen Durchlauf
    removeControl(mapproxy, "heatmap_legend")
    
    if (heatmap != "") {
      hideGroup(mapproxy, "layer_bezirke")
      hideGroup(mapproxy, "layer_stadtteile")
      #---Create heatmap----
      heatmap_polygons <- if(maptype == "Bezirke") {
          geo_bezirke %>%
            mutate(bezirke_join = paste("Bezirk", bezirk)) %>%
            left_join(get_intensity_df(crime_json, heatmap), by = c("bezirke_join" = "stadtteil"))
        } else {
          geo_stadtteile %>%
            left_join(get_intensity_df(crime_json, heatmap), by = "stadtteil")
        }
      pal <- colorNumeric(
        palette = "YlOrRd",
        domain = heatmap_polygons$intensity,
        na.color = "transparent",
        alpha = 1
      )
      addPolygons(mapproxy,
        data = heatmap_polygons,
        layerId = paste("heat_", heatmap_polygons$leaflet_id, sep = ""),
        group = "layer_heatmap",
        label = get_map_layer_name(heatmap_polygons$leaflet_id),
        color = "#003063",
        fillColor = ~pal(heatmap_polygons$intensity),
        weight = 3,
        fillOpacity = 0.8,
        highlightOptions = highlightOptions(
          weight = 6,
          bringToFront = TRUE
        ),
      )
      pal_legend <- colorNumeric(
        palette = "YlOrRd",
        domain = heatmap_polygons$intensity
      )# Legende hinzufügen
      addLegend(mapproxy, 
                pal = pal_legend,
                opacity = 0.8,
                values = heatmap_polygons$intensity, 
                title = paste("Fallzahlen:", heatmap),
                position = "bottomright",
                layerId = "heatmap_legend" # Wichtig zum gezielten Entfernen
      )
      #---------------------
    } else {
      if (maptype == "Bezirke"){
         hideGroup(mapproxy, "layer_stadtteile")
         showGroup(mapproxy, "layer_bezirke")
      } 
      else {
        hideGroup(mapproxy, "layer_bezirke")
        showGroup(mapproxy, "layer_stadtteile")
      }
    }
  })
  
  observeEvent(input$hhmap_shape_click, {
    click_event <- input$hhmap_shape_click
    # Check if an ID was returned (meaning a polygon was clicked)
    if (!is.null(click_event$id)) {
      if(click_event$group == "selected") {
        # Clicked on an already selected area of the map therefore we unselect.
        currently_selected_bezirk("")
        currently_selected_stadtteil("")
        leafletProxy("hhmap") %>%
          clearGroup("selected")
        return()
      }
      
      # The ID of the clicked polygon
      clicked_polygon_id <- click_event$id
      clicked_polygon_id <- sub("^heat_", "", clicked_polygon_id)
      prefix <- get_map_layer_prefix(clicked_polygon_id)
      rest_of_name <- get_map_layer_name(clicked_polygon_id)
      selected_polygon_data <- NULL
      if (prefix == "bez_") {
        currently_selected_bezirk(rest_of_name)
        currently_selected_stadtteil(paste("Bezirk", rest_of_name))
        selected_polygon_data <- geo_bezirke[geo_bezirke[["leaflet_id"]] == clicked_polygon_id,]
      }
      if(prefix == "std_") {
        currently_selected_bezirk(get_bezirk_by_stadtteil(rest_of_name))
        currently_selected_stadtteil(rest_of_name)
        selected_polygon_data <- geo_stadtteile[geo_stadtteile[["leaflet_id"]] == clicked_polygon_id,]
      }
      req(selected_polygon_data)
      #neues Polygon über die anderen legen, wenn ein bezirk angeklickt wurde
      leafletProxy("hhmap") %>%
        clearGroup("selected") %>%
        addPolygons(
          data = selected_polygon_data,
          layerId = id,
          label = selected_polygon_data$bezirk,
          color = "#003063",
          fillOpacity = 0.2,
          weight = 4,
          group = "selected"
        )
    }
  })
  
  output$hhmap <- renderLeaflet({
    leaflet() %>%
      addProviderTiles(providers$CartoDB.Positron) %>%
      addPolygons(
        data = geo_bezirke,
        layerId = ~leaflet_id,
        group = "layer_bezirke",
        label = get_map_layer_name(geo_bezirke$leaflet_id),
        color = "#003063",
        fillOpacity = 0.2,      # Polygon fill transparency
        highlightOptions = highlightOptions(
          fillOpacity = 0.4,
          color = "#003063",
          weight = 4,
          bringToFront = TRUE
        ),
      ) %>%
      addPolygons(
        data = geo_stadtteile,
        group = "layer_stadtteile",
        label = get_map_layer_name(geo_stadtteile$leaflet_id),
        layerId = ~leaflet_id,
        color = "#003063",
        options = pathOptions(pane = "overlayPane"), # Use a leaflet option to ensure it's hidden
        weight = 3,
        fillOpacity = 0.2,      # Polygon fill transparency
        highlightOptions = highlightOptions(
          fillOpacity = 0.4,
          color = "#003063",
          weight = 4,
          bringToFront = TRUE
        ),
      ) %>%
      
      setView(
        lng = 9.98716634776887,
        lat = 53.5488439196432,
        zoom = 11
      )
  })
  output$txt_map_selection_bezirk <- renderUI({
    bez <- currently_selected_bezirk()
    req(bez)
    div(# Überschrift einfügen
      class = "d-flex align-items-baseline gap-2",
      style = "margin-bottom: 2px;",              # Ganz wenig Abstand zum nächsten Element
      tags$b("Bezirk:", style = "font-size: 1.1rem;"), # 'b' ist kompakter als 'h5'
      tags$span(bez, style = "font-size: 1.1rem;")
    )
    
  })
  output$txt_map_selection_stadtteil <- renderUI({
    sdt <- currently_selected_stadtteil()
    req(sdt)
    div(
      class = "d-flex align-items-end gap-2",
      h5(strong("Stadtteil:"), style = "margin-bottom: 0;"),
      p(sdt)
    )
    
  })
  output$grph_top3 <- renderPlotly({
    data_tibble <- map_data_to_top3_plot(
      bezirk = currently_selected_bezirk(),
      stadtteil = currently_selected_stadtteil(),
      year = "2024"
    )
    req(nrow(data_tibble) > 0)
    ggplotly(ggplot(data_tibble, aes(x = reorder(Name, -Erfasst), y = Erfasst, fill = Name)) +
      geom_col(width = 0.7) +
      scale_x_discrete(
        labels = function(x) str_wrap(x, width = 15)) +
      scale_fill_manual(values = c(
        "#b8ecff", 
        "#d5f4ff", 
        "#e6f9ff"
      )) +
      geom_text(
        # Die Text-Ästhetik soll der Wert aus der Spalte 'Erfasst' sein
        aes(label = format(Erfasst, big.mark = ".", decimal.mark = ",")), 
        # Platzierung: Y-Wert des Textes = Wert der Spalte + Offset
        # Wir verwenden den Offset, um den Text knapp über den Balken zu platzieren
        # Wenn Sie den Text IN den Balken setzen möchten, setzen Sie y=Erfasst/2
        vjust = -0.5, # Vertikale Ausrichtung: Negativer Wert platziert Text über dem Punkt
        size = 4,
        fontface = "bold"
      ) +
      labs(
        title = NULL,
        x = "Straftatbestand",
        y = "Anzahl erfasster Fälle"
      ) +
      theme_pander() + #neues theme aus ggthemes packages
      # NEUE ANPASSUNG: Drehen der X-Achsen-Beschriftungen
      theme(
        plot.background = element_rect(
          color = "darkgrey",  # Farbe des Rahmens
          linewidth = 0.4,    # Dicke des Rahmens
          fill = NA         # Füllung: NA = transparent
        ),
        plot.margin = margin(t = 20, r = 20, b = 20, l = 20, unit = "pt"),
        legend.position = "none",
        panel.grid.major.x = element_blank(), 
        panel.grid.minor.x = element_blank(), # vertikale grid lines entfernen
        # X-Achsen-Titel (z.B. "Straftatbestand")
        axis.text.x = element_text(
          angle = 0, 
          size = 9
          ),
        axis.title.x = element_text(
          face = "bold",
          family = "sans",
          # Fügen Sie hier einen Abstand nach OBEN hinzu
          margin = margin(t = 15) # t = top (oben) in Pixeln
        ),
        
        # Y-Achsen-Titel (z.B. "Anzahl erfasster Fälle")
        axis.title.y = element_text(
          face = "bold",
          family = "sans",
          # Fügen Sie hier einen Abstand nach RECHTS hinzu
          margin = margin(r = 15) # r = right (rechts) in Pixeln
        ),
      )
    )
  })
  
  observeEvent(input$vergleich_location, {
    currently_compared_location(c(input$vergleich_location))
  })
  observeEvent(input$vergleich_straftat, {
    currently_compared_crimes(input$vergleich_straftat)
  })
  
  output$vergleich_balkendiagramm <- renderPlotly({
    data_tibble <- map_data_to_plot(
      locations = currently_compared_location(),
      crimes = currently_compared_crimes(),
      year = currently_compared_year()
    )

    ggplotly(ggplot(data_tibble, aes(x = Erfasst, y = reorder(Location, Erfasst, FUN = sum), group = Name, fill = Name)) + #reorder macht die höchsten Werte nach oben, und sortiert nach Gesamtwert
      geom_col(position = position_dodge(width = 0.9)) +
      scale_fill_brewer(palette = "Blues") +
      geom_text(
        # Die Text-Ästhetik soll der Wert aus der Spalte 'Erfasst' sein
        aes(label = format(Erfasst, big.mark = ".", decimal.mark = ",")), 
        # Platzierung: Y-Wert des Textes = Wert der Spalte + Offset
        # Wir verwenden den Offset, um den Text knapp über den Balken zu platzieren
        # Wenn Sie den Text IN den Balken setzen möchten, setzen Sie y=Erfasst/2
        position = position_dodge(width = 0.9),
        hjust = -0.3, #schiebt die Zahl nach rechts
        vjust = 0.5,  #schiebt Zahlen mittig hinter Balken
        size = 4,
        fontface = "bold"
      ) +
      labs(
        title = "Polizeilich registrierte Straftaten",
        x = "Anzahl erfasster Fälle",
        y = NULL
      ) +
      scale_x_continuous(expand = expansion(mult = c(0, 0.15))) +
      theme_classic() + #neues theme aus ggthemes packages
      # NEUE ANPASSUNG: Drehen der X-Achsen-Beschriftungen
      theme(
        plot.background = element_rect(
          color = "darkgrey",  # Farbe des Rahmens
          linewidth = 0.4,    # Dicke des Rahmens
          fill = NA         # Füllung: NA = transparent
        ),
        plot.margin = margin(t = 20, r = 20, b = 20, l = 20, unit = "pt"),
        legend.position = "bottom",
        legend.text = element_text(size = 13),
        legend.title = element_blank(),
        # X-Achsen-Titel (z.B. "Straftatbestand")
        axis.text.x = element_text(
          vjust = 1, 
          hjust = 1,
          size = 12
        ),
        axis.title.x = element_text(
          face = "bold",
          family = "sans",
          # Fügt Abstand nach OBEN hinzu
          margin = margin(t = 15), # t = top (oben) in Pixeln
          size = 12
        ),
        axis.text.y = element_text(
          size = 13
        ),
        # Y-Achsen-Titel (z.B. "Anzahl erfasster Fälle")
        axis.title.y = element_text(
          face = "bold",
          family = "sans",
          # Fügen Sie hier einen Abstand nach RECHTS hinzu
          margin = margin(r = 15) # r = right (rechts) in Pixeln
        )
      ) 
    )
  })
  
}